The Machine Learning Model for Identifying Bogus Profiles in Social Networking Sites

In recent years, social networking sites have gained popularity on the internet, attracting hundreds of thousands of users who spend billions of minutes utilising them. Social networking sites have now taken over many people’s life. Every day, a large number of people create profiles on social media sites and contact others regardless of their location or time. Users’ personal information is at risk on social networking sites posing security risks. It is important to identify the bogus user’s social media identities so that the person who is pushing threats can be uncovered. In this proposed work, the Ensemble learner approach helps to increase the accuracy of detecting fake profiles through automated user profile classification. The study investigates the different text, image, audio, and video attributes that can be used to discern the difference between a fake and a legitimate account. The Ensemble learner approach utilises these attributes and assesses their performance on real-world social media datasets using performance metrics such as accuracy, precision, recall, and F-1 score. The experimental evaluation indicates that the average accuracy achieved by the proposed ensemble learner approach is 97.67%, precision of 92.56% and a recall of 82%, compared to learning models (Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Logistic regression).